（1）首先是序列对应问题：

00: 2011_10_03_drive_0027
01: 2011_10_03_drive_0042
02: 2011_10_03_drive_0034
03: 2011_09_26_drive_0067
04: 2011_09_30_drive_0016
05: 2011_09_30_drive_0018
06: 2011_09_30_drive_0020
07: 2011_09_30_drive_0027
08: 2011_09_30_drive_0028
09: 2011_09_30_drive_0033
10: 2011_09_30_drive_0034

从网上复制的，前面是对应的序列，明确自己需要的序列号。

（2）contrib/kitti_poses_and_timestamps_to_trajectory.py · master · mirrors / michaelgrupp / evo · GitCode

从这个里面下载 kitti_poses_and_timestamps_to_trajectory.py程序

（3）下载对应的poses文件  xx.txt文件

下载链接为（https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_poses.zip）

以及times.txt文件，下载链接为https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_gray.zip

高达20多G。 

 我这里只放了前11个，文件夹结构如下：

 （4）执行如下代码生成tum格式的数据（下面以07文件夹为例）

python2 kitti_poses_and_timestamps_to_trajectory.py data_odometry_poses/dataset/poses/07.txt  sequences/07/times.txt sequences/07/tum_07_gt.txt

或者执行如下脚本

for i in {0..10}
do
	if [ $i -le 9 ]; then
 python2 kitti_poses_and_timestamps_to_trajectory.py data_odometry_poses/dataset/poses/0$i.txt  sequences/0$i/times.txt sequences/0$i/tum_0${i}_gt.txt
	else

 python2 kitti_poses_and_timestamps_to_trajectory.py data_odometry_poses/dataset/poses/$i.txt  sequences/$i/times.txt sequences/$i/tum_${i}_gt.txt
 	fi
done

最终生成的tum数据文件名为tum_xx_gt.txt

evo_traj tum sequences/00/tum_00_gt.txt  -p

evo_traj tum sequences/01/tum_01_gt.txt  -p

evo_traj tum sequences/02/tum_02_gt.txt  -p

evo_traj tum sequences/03/tum_03_gt.txt  -p

 evo_traj tum sequences/04/tum_04_gt.txt  -p

evo_traj tum sequences/05/tum_05_gt.txt  -p

evo_traj tum sequences/06/tum_06_gt.txt  -p

evo_traj tum sequences/07/tum_07_gt.txt  -p

evo_traj tum sequences/08/tum_08_gt.txt  -p

evo_traj tum sequences/09/tum_09_gt.txt  -p

evo_traj tum sequences/10/tum_10_gt.txt  -p

我已经将生成好的数据打包上传了资源，有需要的搜索下载。